[arin-ppml] Revised and Retitled - Draft Policy ARIN-2021-6: Permit IPv4 Leased Addresses for Purposes of Determining Utilization for Future Allocations

Fri Mar 18 15:39:06 EDT 2022

Hi Scott,

I am sorry, I actually penned a long reply to  your initial post but never sent it.

The limit on initial block size is the same as if you came to ARIN seeking a block not for lease, but for your circuit-connected customers. To wit, ARIN will require more than just your statement.

What’s to stop a company from claiming they will be delivering service to connected clients over the next two years, so they should get a pre-approval?

The initial size for a section 8 transfer per 8.5.5 is a /24. To get a larger initial block, evidence is required that should be equivalent in scope and detail regardless of whether the anticipated clients are connected with a circuit or not. Are we seeing such fraud by ISPs to increase their purchase size? 

Finally, if you attest to fraudulent information, you are taking liability risks that also dampen speculation.

Regards,

Mike

PS You ignored my request for evidence of speculation at RIPE where absolutely no needs demonstration has been required for many years. 😉

From: Scott Leibrand <scottleibrand at gmail.com> 
Sent: Friday, March 18, 2022 3:25 PM
To: Mike Burns <mike at iptrading.com>
Cc: Owen DeLong <owen at delong.com>; Andrew Dul <andrew.dul at quark.net>; arin-ppml at arin.net
Subject: Re: [arin-ppml] Revised and Retitled - Draft Policy ARIN-2021-6: Permit IPv4 Leased Addresses for Purposes of Determining Utilization for Future Allocations

On Mar 18, 2022, at 7:37 AM, Mike Burns <mike at iptrading.com <mailto:mike at iptrading.com> > wrote:

Hi Owen, Andrew, and Scott,

Transfer approval of a larger-than-minimum block size requires detailed documentation of the use of at least 50% of the block in 24 months, and that detailed documentation must be officer-attested.  I’m sure we all agree that nobody can approach ARIN for a large initial block without providing believable documentation to ARIN, and the attestation provides actionability against fraud. 

I don’t doubt that believable documentation is required. But my concern, as I stated in the very first reply to this thread that everyone ignored, is:

If you’re going to remove that, what is to stop me from opening a new LIR and stating that I want pre-qualification for a transfer of a /8 to lease out, because I have sales projections that I can lease out a /9 within 24 months, a /10 within 12 months, and a /11 within 6 months? And if I fail to meet my sales projections, I can sell some or all of the /8 after 12 months (presumably at a profit, as prices just keep going up).  

It seems that there should be some limit on initial block size if we’re going to rely exclusively on recipients’ leasing projections instead of requiring use on an operational network.

I take your point about a /8 being infeasible to acquire on the market, but the same point applies at whatever the maximum available size currently is. 

-Scott

Further transfers require proof of utilization of the original transfers. 

This persistent fear of “speculation”, whatever that word means in this context, is belied by the RIPE experience. Will somebody please answer the RIPE experience before bringing up the “speculation” argument?

It’s over 10 years now. The experiment has been performed. We have the data. It’s time to point to evidence instead of holding policy in thrall to assertions of the dangers of speculation.

Remember the biggest damper on speculation is the reality of the market. You can’t just whip up a /8 to be transferred, and if you could, you are looking at spending almost a billion dollars! Do you really think a billion dollar investment in an asset that all the smart people say will be valueless at some point isn’t a damper on speculation? Do you think ARIN would approve an initial transfer of a /8 on the mere promise it will be leased out in two years?

I trust the market and I trust ARIN staff enough to dampen “speculation”. As always, should something damaging appear, we retain the ability to change policy.

Regards,
Mike

From: ARIN-PPML <arin-ppml-bounces at arin.net <mailto:arin-ppml-bounces at arin.net> > On Behalf Of Owen DeLong via ARIN-PPML
Sent: Thursday, March 17, 2022 8:20 PM
To: Andrew Dul <andrew.dul at quark.net <mailto:andrew.dul at quark.net> >
Cc: arin-ppml at arin.net <mailto:arin-ppml at arin.net> 
Subject: Re: [arin-ppml] Revised and Retitled - Draft Policy ARIN-2021-6: Permit IPv4 Leased Addresses for Purposes of Determining Utilization for Future Allocations

I favor the kind of limitations Scott has expressed. I was commenting on the arguments made by Fernando and have not yet had the bandwidth to review the actual policy text in detail.

Owen

On Mar 17, 2022, at 16:17 , Andrew Dul <andrew.dul at quark.net <mailto:andrew.dul at quark.net> > wrote:

The draft policy as currently written does not provide any additional limits against speculation.  As drafted, it allows any organization (including those who do not operate networks) to obtain IPv4 addresses for the purpose of leasing.  

With that policy change what types of limits does the community think would be needed?

Thanks,

Andrew

On 3/17/2022 3:00 PM, Scott Leibrand wrote:

+1 to both Owen and David Farmer's comments. Leasing IPv4 space is likely the best solution for some networks that need those addresses to operate their network. If an organization wants to acquire and lease out IPv4 space without providing bundled IPv4 transit, that should be allowed by policy. It might be useful for ARIN policy to try to slightly dampen speculation by requiring that organizations seeking to acquire large blocks of IPv4 space demonstrate that their current holdings are being efficiently used by the organization they're registered to in whois. I am not sure if this policy proposal does that to my satisfaction, but once we ensure it does so, I would likely support it.

-Scott

On Thu, Mar 17, 2022 at 1:33 PM Owen DeLong via ARIN-PPML <arin-ppml at arin.net <mailto:arin-ppml at arin.net> > wrote:

On Mar 16, 2022, at 15:22 , Fernando Frediani <fhfrediani at gmail.com <mailto:fhfrediani at gmail.com> > wrote:

Hi David

If I understand correctly you seem to have a view that there should be a ARIN policy to permit IPv4 leasing just because it is a reality and we kind of have to accept it in our days. No we don't, and that's for many different reasons.

Well, of course, you are free to deny reality as much as you want. Many people do. It’s not particularly helpful in the discussion, however.

I am used to see people saying the brokers are doing a good thing for the community by facilitating the things which in reality is the opposite. It may look like a good things, but the real beneficiaries are only them who profit from it without much concern of what is fair or not to most organizations involved.

You are actually mistaken here. I used to think as you do, actually. I was very resistant to the first “specified transfer” policies because of some of the reasons you describe. However, what you are failing to recognize is that:

+ Brokers and specified transfers were going to happen with or without the RIRs. If they happened without the RIRs,

there’d be no accurate record of who was using which address space and the provenance of addresses would be

very difficult to support or defend.

* Benefit to the community from brokers: (ethical) brokers are familiar with the rules in the RIRs in which

they operate and can assist their customers in accurate and compliant registration updates and

aid in keeping the allocation database(s) accurate.

+ With the economic realities of IPv4 addresses becoming progressively more and more expensive and the advent

of ISPs with limited IPv4 resources available, it is inevitable that more and more IP service providers will be

doing one or more of the following:

+ Separate surcharges for IPv4 addresses

+ Expecting customers to supply their own IPv4 addresses

+ Surcharges for IPv4 services

+ IPv4 “installation charges” large enough to cover the procurement of addresses

* Brokers assist ISPs and customers in many of the above circumstances.

+ With a variety of organizations holding IPv4 addresses that may or may not even known they have them and whose

IPv4 resources may vastly exceed their needs, it is (arguably) desirable to have those addresses be transferred to parties

that have current need for IPv4 addresses.

* Brokers provide a valuable service to the community identifying and marketing these resources

* Paid transfers provide an incentive for entities to make more efficient use of the resources they have in order

to monetize the resources they no longer need. Brokers are frequently able to assist in this process.

+ With the high cost of acquisition, IPv4 addresses have become a capital intensive part of any network-dependent

business model that must support IPv4. Further, there is some risk that this capital outlay may be fore a resource

which will abruptly and quickly lose its value and no longer be needed well before it can be amortized as a capital

expenditure. As such, it may make sense for some entities to transfer that risk to another organization by using

a lease structure instead of purchasing the addresses outright.

* Brokers that provide IPv4 leasing in an ethical and policy compliant way provide a valuable service

to these businesses. Yes, their price per address may eventually be more than it would have cost

them to purchase the addresses, but the same is true of virtually any rental situation.  On the other hand,

that excess helps offset the risk that the lessor is taking by owning a resource that may or may not remain

valuable and may or may not continue to produce revenue.

IP Leasing is very different from IP Transfer which I see not problem they continue doing it. IP Transfer at least we have some guarantees that the directly receiving organization really justify for them and that is a quiet important (I would say fundamental) point to look at, because that is fairer to everyone involved. What guarantees we have when a IP Leasing is done in that sense, that fairness start to lack here.

If we set the policies up correctly, we should have the same exact guarantees on a lease.

If $ISP acquires a /10 through transfer and then issues various subordinate prefixes to their customer, the only guarantee

you have that $ISP’s customers who receive the addresses really justify them is that $ISP says so. We generally trust $ISP

to act in good faith.

If $LESSOR acquires a /10 through transfer and then leases various subordinate prefixes to their customers, we have pretty

much the same guarantee with the additional bit that $CUSTOMER is at least willing to pay enough for the addresses to $LESSOR

to make the lease make sense. In general, I think it is somewhat safe to assume that $CUSTOMER is not going to make a

monthly recurring payment to $LESSOR for something they don’t intend to use. If one’s intent is to deprive the market and

inflate the price, then the risk profile for such a transaction is vastly more favorable if you purchase rather than lease.

Sure, there could be lessors that don’t get reasonable justification for allocations from their customers, but there are most

certainly ISPs in that category as well. Either way, you’ve got very little assurance. A lessor can provide just as much

justification to an RIR for the addresses they will allocate to leases as an ISP can for addresses they will lease to their

customers. The only difference is a lease with connectivity from the same company or a lease from a company other than

the one(s) providing connectivity.

People see the brokers are doing a favor to organizations in general by facilitating they get some chunks of IPv4, but that in reality makes the cost of IPv4 for both leasing and transfer more and more expensive as it makes organization even more dependent from these those crumbs that seem to be offered with good intention but in reality it is feeding a system that is contrary the interests to most organizations involved.

Just as you are free to mount, balance, and rotate your own tires, or, you can go to a tire store and have them perform that service for a fee, brokers provide a service for a fee. If you want to obtain addresses in the transfer market without a broker, you’re still free to do that. Brokers are not driving the cost of IPv4… The scarcity and difficulty of operating with IPv4 is driving the cost of IPv4. Brokers are along for the ride providing a service and collecting a fee for that service. Whether that fee is reasonable or not is (and should be) entirely in the eye of the customer. Customers are always free to walk away and find a different supplier or look for their addresses independently.

It may sound a cliche but IPv4 is over and organizations must learn how to survive with what they have, reinvent themselves and make better used of their IPv4 resources, deploy a proper CGNAT, deploy IPv6 either they like it or not, etc. If an organization have so little or none and need some minimal amount is fine they seek for a Transfer of a minimal amount with the help of brokers. 

I agree. However, the increasing cost of IPv4 is a natural and organic part of that process and sticking our heads in the sand and pretending that it is not the economic reality of how the current world works will not help anyone. Not the community, not organizations that are short on IPv4 resources, and not the RIRs who are only useful so long as their databases provide a reasonably accurate reflection of the actual utilization of the address space and who controls it.

A broker is an LIR just like an ISP. Since ISPs are now charging for addresses independent of connectivity and bandwidth, it only makes sense that customers can shop for them separately from different suppliers. Just like you can buy tires for your car from the dealership or from some other store that sells and supports tires, IPv4 addresses are moving that way as well. The RIRs can either recognize this and adapt to it with policies that make sense and preserve some of the things you’ve outlined as concerns above, or, they can simply deny the reality of IPv4 leasing and lose track of how addresses are actually managed for some significant chunks of the internet.

Encouraging IP Leasing as if it were something normal just "because it exists today" is a shot in the foot that in the long term only worsens the existing scenario, it feeds a market without much discretion increasing final prices for everyone and what is the worst of all, creates even more unfairness for everyone who has always submitted to the rules we have until today for distributing addresses to those who really have a real justification to keep control of that resource that does not belong to them.

I don’t believe that a policy that merely allows IPv4 leasing can be said to encourage it. Rather, it permits it, recognizes that it exists and is not going to stop existing just because policy pretends it can’t exist.

The market is not likely to be significantly swayed by policy in terms of pricing, with the exception that AFRINIC has been able to preserve a devalued price on addresses within their region due to their restrictive lack of a transfer policy for moving addresses to/from AFRINIC. However, while this has the effect of keeping AFRINIC IPv4 addresses less expensive on the open market, it also leads to a significant amount of utilization of those addresses outside of policy and quite a bit of hoarding of addresses by some of AFRINIC’s largest members. ARIN’s counsel has advised against naming names here, so I won’t, but if you want names, contact me off list.

Owen

Regards
Fernando

On 16/03/2022 13:09, David Farmer via ARIN-PPML wrote:

Yes, bundling IPv4 addresses with bandwidth is permitted, and in the past was common practice, heck even the expected practice. However, the fact that IPv4 address demand isn't decreasing significantly, the costs to acquire new IPv4 addresses are increasing significantly, and with the increasing commoditization of bandwidth, it is no longer economically viable to bundle bandwidth, and its associated connectivity, with IPv4 addressing. This is driving a structural separation of bandwidth, connectivity, and IPv4 addressing, from each other, instead of bundling them together as in the past.

Let me state that differently; ISPs are being driven, buy cost conscience consumers, to separate the costs of bandwidth and the costs of the IPv4 addresses needed to utilize the bandwidth from each other.  Minimally this separation is achieved by accounting for the costs on separate line items of a common bill from a single provider. However, price competition for bandwidth and IPv4 addresses separately will inevitably drive a structural separation between the two. Consumers will want the best price they can get for bandwidth and the best price they can get for IPv4 addresses, regardless of whether they come from a single provider or not.

Some may argue this is being driven by the existence of address brokers, and their desire to make money, I disagree. While address brokers making money is the grease that keeps this machine working, the need for the machine is driven by; IPv4 free pool exhaustion, the increasing cost of IPv4 addresses, and the lack of adoption of IPv6.

In other words, address brokers wouldn't exist if there wasn't a demand for their services.

In short, the economic conditions that allowed for and even encouraged the bundling of IPv4 addresses with bandwidth and connectivity no longer exist, that world is gone. While I have not personally yet determined if I support this particular policy text, nevertheless, the time has come to recognize the next step in this inextricable evolution of IPv4 address policy by the ARIN policy community and permit IPv4 leasing.

Thanks.

On Fri, Mar 11, 2022 at 5:05 PM John Santos <john at egh.com <mailto:john at egh.com> > wrote:

I disagree.  The addresses are useless unless they ALSO purchase access and 
routing from another network operator.  How is this cheaper?

It is and always has been allowed to lease bundled access of addresses and 
connectivity from a LIR, without any expense for purchasing those addresses.

On 3/11/2022 12:13 PM, Tom Fantacone wrote:
> I support the proposal as written.
> 
> It facilitates the provision of a valuable service to a large swath of the ARIN 
> community, namely the ability of network operators with an operational need to 
> lease IPv4 addresses from 3rd party lessors at a fraction of the cost of 
> purchasing those addresses.  Too often we have seen network operators justify 
> their need for IPv4 space only to find that they can't afford to make the 
> purchase.  They end up using CGNAT or some other sub-optimal solution.
> 
> Bill, regarding your point "B", by providing IPv4 leasing, these 3rd parties are 
> certainly performing a function that ARIN does not.
> 
> 
> 
> ---- On Thu, 10 Mar 2022 17:46:36 -0500 *William Herrin <bill at herrin.us <mailto:bill at herrin.us> >* wrote ----
> 
>     On Wed, Mar 9, 2022 at 8:24 PM ARIN <info at arin.net <mailto:info at arin.net>  <mailto:info at arin.net <mailto:info at arin.net> >>
>     wrote:
>      > * ARIN-2021-6: Permit IPv4 Leased Addresses for Purposes of Determining
>     Utilization for Future Allocations
> 
>     I continue to OPPOSE this proposal because:
> 
>     A) It asks ARIN to facilitate blatant and unapologetic rent-seeking
>     behavior with changes to public policy.
> 
>     B) It proposes that third parties perform precisely and only the
>     functions that ARIN itself performs without any credible compliance
>     mechanism to assure the third party performs to ARIN's standards or in
>     accordance with the community's established number policy.
> 
>     Regards,
>     Bill Herrin
> 
> 
>     -- 
>     William Herrin
>     bill at herrin.us <mailto:bill at herrin.us>  <mailto:bill at herrin.us <mailto:bill at herrin.us> >
>     https://bill.herrin.us/ <https://bill.herrin.us/>
>     _______________________________________________
>     ARIN-PPML
>     You are receiving this message because you are subscribed to
>     the ARIN Public Policy Mailing List (ARIN-PPML at arin.net <mailto:ARIN-PPML at arin.net> 
>     <mailto:ARIN-PPML at arin.net <mailto:ARIN-PPML at arin.net> >).
>     Unsubscribe or manage your mailing list subscription at:
>     https://lists.arin.net/mailman/listinfo/arin-ppml
>     <https://lists.arin.net/mailman/listinfo/arin-ppml>
>     Please contact info at arin.net <mailto:info at arin.net>  <mailto:info at arin.net <mailto:info at arin.net> > if you experience any
>     issues.
> 
> 
> 
> 
> _______________________________________________
> ARIN-PPML
> You are receiving this message because you are subscribed to
> the ARIN Public Policy Mailing List (ARIN-PPML at arin.net <mailto:ARIN-PPML at arin.net> ).
> Unsubscribe or manage your mailing list subscription at:
> https://lists.arin.net/mailman/listinfo/arin-ppml
> Please contact info at arin.net <mailto:info at arin.net>  if you experience any issues.

-- 
John Santos
Evans Griffiths & Hart, Inc.
781-861-0670 ext 539
_______________________________________________
ARIN-PPML
You are receiving this message because you are subscribed to
the ARIN Public Policy Mailing List (ARIN-PPML at arin.net <mailto:ARIN-PPML at arin.net> ).
Unsubscribe or manage your mailing list subscription at:
https://lists.arin.net/mailman/listinfo/arin-ppml
Please contact info at arin.net <mailto:info at arin.net>  if you experience any issues.

-- 

===============================================
David Farmer               Email:farmer at umn.edu <mailto:Email%3Afarmer at umn.edu> 
Networking & Telecommunication Services
Office of Information Technology
University of Minnesota   
2218 University Ave SE        Phone: 612-626-0815
Minneapolis, MN 55414-3029   Cell: 612-812-9952
=============================================== 

_______________________________________________
ARIN-PPML
You are receiving this message because you are subscribed to
the ARIN Public Policy Mailing List (ARIN-PPML at arin.net <mailto:ARIN-PPML at arin.net> ).
Unsubscribe or manage your mailing list subscription at:
https://lists.arin.net/mailman/listinfo/arin-ppml
Please contact info at arin.net <mailto:info at arin.net>  if you experience any issues.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.arin.net/pipermail/arin-ppml/attachments/20220318/dbf5f652/attachment.htm>